Using Drones to Map and Model Flood Risks in Dar es Salaam, Tanzania

The first in our series of cases studies in the use of drones in flood mapping and flood disaster risk reduction and recovery looks at how drones were used to map and model flood risks in Dar es Salaam, Tanzania.

Dar es Salaam is one of Africa’s fastest growing cities. Over 70 per cent of the population lives in informal, unplanned settlements with poor infrastructure, extremely vulnerable to the impacts of severe weather. The city regularly experiences extreme rainfall events, often resulting in severe flooding and devastation. Last year around 20 people were killed and thousands left homeless after two flood events (March and May) struck within the space of a few weeks.

Figure 1 Extract orthomosaic
Figure 1 Extract orthomosaic

Case Study: Mapping – Flood Mapping for Disaster Risk Reduction: Obtaining High-Resolution Imagery to Map and Model Flood Risks in Dar es Salaam by Denise Soesilo for Swiss Foundation for Mine Action (FSD), as part of the Drones in Humanitarian Action project.

Details

  • Type of system: eBee Fixed Wing Mapping Microdrone
  • Deploying Agency: World Bank and Humanitarian OpenStreetMap Team
  • Piloting Agency: Drone Adventures
  • Dates of Deployment: February and March 2015

Summary

Drones provided high-resolution, up-to-date imagery that was essential to the development of exposure maps and for modeling flood risks in Dar es Salaam. The success of the project has led to further applications and to government funding for infrastructure improvements in vulnerable areas identified by the project.

Background

With an estimated population of about 5 million and an annual growth rate of 8 per cent, Dar es Salaam, Tanzania, is Africa’s fastest growing city.[1] Over 70 per cent of the people live in informal, unplanned settlements with inadequate infrastructure. In addition, heavy rainfalls twice a year result in significant flood risks. The objectives in Dar es Salaam were to 1) obtain high quality and up to date exposure maps of affected communities and as a stretch goal 2) to create a hydrological model using elevation data. Both of these would form important elements to managing flood risks but neither had been previously available due to a lack of high quality digital imagery. In response to this data gap, the World Bank and the Red Cross began to support and scale up community mapping. A coalition comprising of the City Council of Dar es Salaam, the Tanzania Commission for Science and Technology (COSTECH), the University of Dar es Salaam, Ardhi University and Buni Innovation Hub formed a project in early 2015. Called Ramani Huria – Swahili for “Open Mapping” – the coalition set out to map flood-prone areas in Dar es Salaam, to make the mapping data openly available and to promote the use of this information in government decisions to mitigate flooding”[2]

Implementation

The data collection was done in two phases both times using an eBee fully automated mapping drone manufactured by Sensefly. In the first phase, Drone Adventures mapped a section of several kilometers of the highway, the river and part of the Tandale neighborhood. The methods and results were shared with local government officials and city planners, who then provided input for priority areas for the second mapping phase.[3] During the second phase the majority of the data was obtained, including 20 000 optical images covering an area of 88 km2 with an average resolution of about 5 cm.[4] In order to achieve this kind of coverage, 2-3 eBee drones were flying simultaneously operated by one pilot. Drone Adventures used eBee Image processing software, Postflight Terra 3D to patch the separate images together to create two main datasets: an optical 2D orthomosaic model and a 3D surface model. The processing was done in Switzerland over a period of 6 weeks with high-memory speed processors that produced a total of 700 GB of data. The dataset was so large that it had to be mailed physically on a hard drive to the Tanzania World Bank office.

Figure 4 Level of detail available on OpenStreetMap shows the details of houses and roads. Retrieved on 3 December 2015. Source: http://www.openstreetmap.org/
Figure 4 Level of detail available on OpenStreetMap shows the details of houses and roads. Retrieved on 3 December 2015. Source: Open Street Map
Figure 3 Google maps snapshot shows little information on what are actually densely populated areas of Dar es Salaam. Retrieved on 3 December 2015.
Figure 3 Google maps snapshot shows little information on what are actually densely populated areas of Dar es Salaam. Retrieved on 3 December 2015.

Data Use and Information Products

COSTECH has made the data available to the public and has published raw images on OpenAerialMap, a collection of openly licensed aerial imagery. Several information products have already emerged. Upon obtaining the 2D orthomosaic imagery, The Humanitarian OpenStreetMap Team (HOT) organized local volunteers to fist determine the respective communities’ mapping priorities and to then digitize roof prints, roads, drains, and trees and to mark businesses and roads. Through this exercise the aerial imagery was converted into digital street maps.[5] With this information houses, businesses, and critical infrastructure have been mapped with unprecedented accuracy. Figures 3 and 4 were available maps of the same area before and after the drone imaging. Figure 5 shows a detailed map of the districts drainage system that was made in the community mapping initiative.

Evaluation

Impact

It is too early to understand the full impact that having this kind of high quality imagery will have on decision-making and flood risk mitigation. However a few clear uses and advantages have already emerged. The Tandale Ward Office and the Department of Urban Planning of the Kinondoni Municipal Council are now using the maps for urban planning purposes. Local authorities see the maps as starting points in the evaluation of areas subject to improvements, and for the response to flooding and related disease outbreaks. The detailed ward maps allow public health officials to identify specific houses with known cholera patients in case of an outbreak allowing for much more efficient response through targeted dispatch of medical care. The exposure maps that include up-to-date buildings and infrastructure now enable planners to identify places that get flooded and to pinpoint where drainage systems need to be installed.[7] This new capacity has even helped the city council acquire funding from the central government to install drainage systems in vulnerable places. According to Mark Iliffe from the World Bank, “creating a flood inundation model is the holy grail but we have not been able to achieve this yet”[8] . However, the work is still in progress. Deterministic flood models have been created with the drone images that can help understand potential inundating zones.

Choice of Method

“You can far more reliably launch a drone and get imagery compared to a satellite or plane because you can fly underneath the cloud cover”. –Edward Anderson

In the case of modeling flood risk and obtaining exposure maps for Dar es Salaam, having high resolution and up-todate imagery was seen by a variety of involved actors as highly valuable. Conventional methods that rely on qualitative data and satellite imagery with a resolution of 30 cm were available to complete this mapping,[9] but were severely limited. The city is growing and changing at such a rate that satellite imagery often becomes outdated within a year. Aerial images obtained with airplanes were often limited due to cloud cover.

Community Engagement

Local authorities and communities were involved throughout the planning, data collection and data use phases. In particular, COSTECH, the technology branch of the Ministry of Communication, Science and Technology, helped decide which districts needed to be mapped, and Ministry officials were present every day of the drone flights. As a result of the transfer of knowledge and the Ministry’s acquisition of two drones, COSTECH now has the capacity to conduct drone flights and to process flight data.

Social Acceptance

Often little is known about community social acceptance on the use of drones for humanitarian or development purposes. However, in Dar es Salaam FHI 360 interviewed 14 high-level government officials and 208 community members who witnessed drone flights over the city. The study found that both “community witnesses and government officials were positive about the potential of [drone] technology in Tanzania”.[10] Among many other potential uses, both community members and government officials noted disaster relief and the transport of medical supplies as promising applications. General concerns included the potential for accidents and for threats to national security. In addition, some of the latrines in the mapped communities do not have covers, raising concerns related to visual privacy among community members.

Figure 5 Detailed drainage map that was drawn up with help of the orthomosaics.
Figure 5 Detailed drainage map that was drawn up with help of the orthomosaics.

Plans for Adaptation

Both the Tanzanian government and the World Bank plan on continuing the use of drones for mapping purposes. The government is seeing the maps as a baseline to understand how the city is developing and changing following improvements to infrastructure and COSTEC has now developed the capacity to do follow-up mapping on their own. On the strength of the success of the initial mapping, the Tanzanian government is planning to complete mapping of the remaining areas in Dar es Salaam, as well as creating annual maps for change detection. The World Bank also intends to continue to use drones, however some concerns regarding the future regulatory environment have been expressed. This effort was only possible because of the close collaboration with the Tanzanian government and because there are no limiting regulations on the use of drones in place

Resources

Open Aerial Map – Repository of aerial imagery that was collected with the drones:
http://beta.openaerialmap.org/#/39.287109375,-6.926426847059551,9?_k=dnojj1

Open Street Map – The detailed city map of Dar es Salaam.
http://www.openstreetmap.org/#map=16/-6.7918/39.2398

Ramani Huria – Mapping Initiative in Dar es Salaam
http://ramanihuria.org/

Author: Denise Soesilo

With special thanks to: Edward Anderson (World Bank), Paul Uithol (HOTOSM), Mark IIiffe (World Bank), Otto Simonett (Zoi Environment Network), Joel Kaiser (Medair), Audrey Lessard-Fontaine (CartONG), Martin Noblecourt (CartONG),for their inputs and comments.

[1] African Development Bank Group. Tracking Africa’s Progress in Figures. 2014
[2] “Mark’s Brain”. Personal Blog of Mark Iliffe.
[3] Edward Anderson. World Bank. Interview from 13 November 2015.
[4] David Rovira, Drone Adventures. Interview from 28 October 2015.
[5] Humanitarian OpenStreetMap http://tasks.hotosm.org/project/1120
[6] Edward Anderson. World Bank. Interview from 13 November 2015.
[7] Osiligi Lossai Tandale Ward Executive Officer and Juliana Letara Head of Urban Planning Kinondoni Municipal Council. Interviewed by Mark Iliffe. 21 October 2015.
[8] Mark Iliffe. World Bank. Interview from 19 October 2015
[9] Mark Iliffe. World Bank. Interview from 19 October 2015
[10] Eichleay M, Mercer S, Murashani J, Evens E. Using Unmanned Aerial Vehicles for Development: Perspectives from Citizens and Government Officials in Tanzania. FHI 360. 2016. Published on ICTworks January 2016 (pdf)


The above article was originally published by the Swiss Foundation for Mine Action (FSD) and is re-published here with permission. Further information is available here (pdf)